It's speculation, but DMT probably has a molecular structure that's sufficiently comaptible with the crystal structure of quartz for it to mobilize into the quartz and distribute interstitially as crystallographic defects. The heat provides the energy of activation necesary for the quartz's vibrations to let individual molecules enter its matrix and from there they're driven deeper by a concentration gradient.
These particles would then be somewhat regularly spaced by the crystal matrix of the quartz (not every compatible interstitial spot occupied, but higher concentration near the contact source and more finely spread deeper into the crystal).
Reheating would not remove or significantly move them since they are at an extremely favorable point in terms of Gibbs Free Energy once they settle into the quartz matrix. I would expect to see some "evening out" if heated for prolonged periods, but there would be no real driving force to move/remove them and it wouldn't be enough to destroy them since the quartz crystal lattice at that point would be helping hold the DMT molecule in shape preventing degredation.
This would naturally cause a color change if I'm right about what's happening.
Please take all of this speculation with a grain of salt. It's been a while since I've studied crystal systems. I'm also waffling on whether GFE is the right notion to explain why it's such a favorable energy state once they settle into the matrix. It's a combination of entropy, enthalpy, and surface energy at the interstitial site. I should have saved some of those text books from my youth...
Incidentally, I'd love to see a picture of your nail if you're inclined to share. Just be sure to strip the meta-data before you upload it anywhere.
